FiniMOM

FiniMOM performs Bayesian fine-mapping to detect and identify independent causal variants within GWAS loci by modeling non-null effects and accounting for linkage disequilibrium patterns.

Key Features:

• Bayesian fine-mapping: Implements a Bayesian fine-mapping framework for identification of causal variants in GWAS loci.
• Nonlocal inverse-moment prior: Employs a nonlocal inverse-moment prior tailored to model non-null effects effectively in finite sample sizes.
• Beta-binomial prior for causal count: Incorporates a beta-binomial prior to estimate the number of causal variants with adjustable parameters.
• Linkage disequilibrium handling: Accounts for linkage disequilibrium patterns and provides parameter adjustments to mitigate LD reference misspecification.
• Multiple causal variant modeling: Explicitly models loci with multiple causal variants to improve detection in complex loci.
• Improved credible set coverage and power: Demonstrates superior credible set coverage and power compared to SuSiE (Summarized data-based Set Identifiability and Estimation) in reported simulations.
• Validation in molecular trait simulations: Validated through simulation studies replicating GWAS scenarios focused on circulating protein levels.

Scientific Applications:

• GWAS fine-mapping: Pinpointing independent causal variants within regions identified by genome-wide association studies.
• Genetic architecture dissection: Investigating loci with multiple causal variants to elucidate the genetic basis of complex traits and diseases.
• Molecular phenotype analysis: Applied to circulating protein levels GWAS and similar molecular trait studies.

Methodology:

Uses a Bayesian framework with a nonlocal inverse-moment prior for non-null effects and a beta-binomial prior to estimate the number of causal variants, with adjustable parameters to mitigate linkage disequilibrium reference misspecification; performance was evaluated via simulation studies replicating GWAS scenarios focused on circulating protein levels and benchmarked against SuSiE.